{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2RWJQ3AADVGGQHLFT2WY2AM3V5","short_pith_number":"pith:2RWJQ3AA","schema_version":"1.0","canonical_sha256":"d46c986c001d4c681d659ead8d019baf5c1da62f9384d4b552d2ba53b7f2cf03","source":{"kind":"arxiv","id":"2502.08106","version":3},"attestation_state":"computed","paper":{"title":"PoGDiff: Product-of-Gaussians Diffusion Models for Imbalanced Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Hao Wang, Sizhe Wei, Xiaoming Huo, Ziyan Wang","submitted_at":"2025-02-12T04:07:14Z","abstract_excerpt":"Diffusion models have made significant advancements in recent years. However, their performance often deteriorates when trained or fine-tuned on imbalanced datasets. This degradation is largely due to the disproportionate representation of majority and minority data in image-text pairs. In this paper, we propose a general fine-tuning approach, dubbed PoGDiff, to address this challenge. Rather than directly minimizing the KL divergence between the predicted and ground-truth distributions, PoGDiff replaces the ground-truth distribution with a Product of Gaussians (PoG), which is constructed by c"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2502.08106","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-12T04:07:14Z","cross_cats_sorted":["cs.AI","cs.CV","stat.ML"],"title_canon_sha256":"6544184e032e8543e52124285a01a861a0e091b21fe75806fc18a57ca3e149cf","abstract_canon_sha256":"34ab4a7f202e11bd1f0b5d816b24db8c138128761959eb76406ce127d081ed65"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:21:42.096575Z","signature_b64":"RR8symkCVi6VQ0gBacWdxiesMTtZnelqLmvsJMzOqRroFVLD3wkECXfFGfddGUsixu8yJeT94Jim/bLaHDe/CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d46c986c001d4c681d659ead8d019baf5c1da62f9384d4b552d2ba53b7f2cf03","last_reissued_at":"2026-07-05T11:21:42.096096Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:21:42.096096Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PoGDiff: Product-of-Gaussians Diffusion Models for Imbalanced Text-to-Image Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Hao Wang, Sizhe Wei, Xiaoming Huo, Ziyan Wang","submitted_at":"2025-02-12T04:07:14Z","abstract_excerpt":"Diffusion models have made significant advancements in recent years. However, their performance often deteriorates when trained or fine-tuned on imbalanced datasets. This degradation is largely due to the disproportionate representation of majority and minority data in image-text pairs. In this paper, we propose a general fine-tuning approach, dubbed PoGDiff, to address this challenge. Rather than directly minimizing the KL divergence between the predicted and ground-truth distributions, PoGDiff replaces the ground-truth distribution with a Product of Gaussians (PoG), which is constructed by c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.08106","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2502.08106/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2502.08106","created_at":"2026-07-05T11:21:42.096158+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.08106v3","created_at":"2026-07-05T11:21:42.096158+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.08106","created_at":"2026-07-05T11:21:42.096158+00:00"},{"alias_kind":"pith_short_12","alias_value":"2RWJQ3AADVGG","created_at":"2026-07-05T11:21:42.096158+00:00"},{"alias_kind":"pith_short_16","alias_value":"2RWJQ3AADVGGQHLF","created_at":"2026-07-05T11:21:42.096158+00:00"},{"alias_kind":"pith_short_8","alias_value":"2RWJQ3AA","created_at":"2026-07-05T11:21:42.096158+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.08960","citing_title":"A Comprehensive Survey on Imbalanced Data Learning","ref_index":37,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5","json":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5.json","graph_json":"https://pith.science/api/pith-number/2RWJQ3AADVGGQHLFT2WY2AM3V5/graph.json","events_json":"https://pith.science/api/pith-number/2RWJQ3AADVGGQHLFT2WY2AM3V5/events.json","paper":"https://pith.science/paper/2RWJQ3AA"},"agent_actions":{"view_html":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5","download_json":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5.json","view_paper":"https://pith.science/paper/2RWJQ3AA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.08106&json=true","fetch_graph":"https://pith.science/api/pith-number/2RWJQ3AADVGGQHLFT2WY2AM3V5/graph.json","fetch_events":"https://pith.science/api/pith-number/2RWJQ3AADVGGQHLFT2WY2AM3V5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5/action/storage_attestation","attest_author":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5/action/author_attestation","sign_citation":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5/action/citation_signature","submit_replication":"https://pith.science/pith/2RWJQ3AADVGGQHLFT2WY2AM3V5/action/replication_record"}},"created_at":"2026-07-05T11:21:42.096158+00:00","updated_at":"2026-07-05T11:21:42.096158+00:00"}